xtensor.random Random number generator operations

Creating RNG variables

pytensor.xtensor.random.variable.rng(name=None)

Create a symbolic xtensor random number generator variable.

This creates a root variable with no data attached, suitable for use as a function input. When compiling a function, use pytensor.In(rng, mutable=True) to allow in-place RNG updates.

Parameters:

name (str, optional) – Name for the variable.

Returns:

A symbolic xtensor RNG variable.

Return type:

XRandomGeneratorVariable

Examples

>>> import numpy as np
>>> import pytensor
>>> import pytensor.xtensor.random as pxr

>>> rng = pxr.rng("rng")
>>> next_rng, x = rng.normal(0, 1, extra_dims={"a": 3})
>>> fn = pytensor.function([pytensor.In(rng, mutable=True)], [next_rng, x])
>>> rng_val = np.random.default_rng(153)
>>> rng_val, draw = fn(rng_val)
>>> draw.shape
(3,)

pytensor.xtensor.random.variable.shared_rng(value=None, *, seed=<object object>, name=None, borrow=False)

Create a shared xtensor random number generator variable.

The RNG state is stored internally and can be updated across function calls via the updates parameter of pytensor.function.

Parameters:

• value (numpy.random.Generator, optional) – The initial RNG state. If None, a new numpy.random.default_rng(seed) is used.

• seed (int, optional) – Seed for the default RNG. Only used when value is None.

• name (str, optional) – Name for the shared variable.

• borrow (bool) – If True, use the provided value directly without copying.

Return type:

XRandomGeneratorSharedVariable

Examples

>>> import numpy as np
>>> import pytensor
>>> import pytensor.xtensor.random as pxr

Create from an existing Generator:

>>> rng = pxr.shared_rng(np.random.default_rng(153), name="rng")
>>> next_rng, x = rng.normal(0, 1)
>>> fn = pytensor.function([], x, updates={rng: next_rng})
>>> fn()
array(1.45769255)

Or create from a seed directly:

>>> rng2 = pxr.shared_rng(seed=153, name="rng2")
>>> next_rng2, x2 = rng2.normal(0, 1)
>>> fn2 = pytensor.function([], x2, updates={rng2: next_rng2})
>>> fn2()
array(1.45769255)

Use set_value to reset the RNG state:

>>> rng.set_value(np.random.default_rng(153))
>>> fn()
array(1.45769255)
>>> rng.set_value(seed=153)
>>> fn()
array(1.45769255)

Distributions

All distributions are available as methods on XRandomGeneratorVariable (e.g. rng.normal()). They accept core_dims and extra_dims parameters for named-dimension support.

class pytensor.xtensor.random.variable.XRandomGeneratorVariable(type, owner, index=None, name=None)

The Variable type used for xtensor random number generator states.